Key takeaways
- The knowledge most at risk in construction is judgment, not procedure: how a veteran reads a job heading for trouble before anything has visibly gone wrong. The codebook is already written down; that instinct is not.
- About 41% of the construction workforce is projected to retire by 2031, and the majority of the workers the industry needs through 2027 are required just to backfill those retirements, not to grow. The experience is leaving faster than it can be replaced.
- Capture the highest-judgment, hardest-to-replace knowledge first, starting with whoever is closest to leaving. You cannot sit down with every veteran at once, so triage is the whole game.
- A tradesperson or superintendent’s realistic role is to demonstrate, explain, and confirm accuracy, not to build the course. Asking your best people to do instructional design is the fastest way to lose their time and their cooperation.
- Knowledge outlasts the worker only if you capture before the retirement date. Reconstructing a skill after the expert has already gone is slower, more expensive, and sometimes not possible at all.
The pour that sets up wrong. A sequence that slips a week because two trades got stacked on top of each other. The weld that comes back at inspection. Most of what keeps those from happening lives in the head of someone who has been doing the work for thirty years, and in construction, a lot of those people are within a few years of retiring.
Few problems give you this much warning. The retirements are visible years out, and the industry still tends to arrive unready. The reason is not that the stakes are unclear. It is that the knowledge is rarely written down anywhere useful, the people who hold it are busy holding production together, and the usual fix, asking them to write down what they know, almost never produces what you actually need.
This piece is about the mechanics of getting it out: what to capture first when you only have a few veterans left, what a capture process looks like on a working jobsite, and how to scope it so the work gets finished without pulling your best superintendent or your best hand off the job for a month. Whether you run your own crews or manage a roster of subs, the knowledge at risk looks different, and so does what you do about it. The aim is a process you could run yourself, with or without a partner.
How big is the retirement gap in construction?
It is large, and it is closer than most workforce plans assume. The National Center for Construction Education and Research projects that about 41% of the current construction workforce will retire by 2031, and the replacement pipeline cannot move that fast. Apprenticeships take years to turn out a fully skilled worker, and the trades skew older every year.
The demand side tells the same story. Associated Builders and Contractors projects the industry needs roughly 349,000 net new workers in 2026 and 456,000 in 2027, and a majority of that demand is just to backfill retirements rather than to staff new work. Deloitte estimates the industry could lose close to $124 billion in output if the gap persists, and reports that only about 10% of current workers are under 25.
The figures undercount the real loss, because a headcount does not show what each of those people carried. When a long-tenured superintendent retires, you lose a set of judgments that kept jobs on schedule and out of trouble. When a master tradesperson retires, you lose technique that was never on paper.
What knowledge is worth capturing first?
Start where the knowledge is hardest to replace and closest to walking out the door. You will never capture everything, and you rarely get more than a few sessions with the people who matter most, so the first real decision is triage.
The codebook, the spec, the manufacturer’s instructions: those are already written down, and they are not what you are at risk of losing. What is at risk is the layer on top of them. The foreman who senses a trench is wrong before the engineer arrives. The estimator whose number is right because of two hundred jobs’ worth of pattern recognition. The superintendent who sequences three trades so they are never standing on each other. None of that sits in a manual, and most of it the person no longer notices they know.
A simple way to sort it:
| Knowledge at risk | How hard to replace | Risk it leaves soon | Capture priority |
|---|---|---|---|
| A retiring expert’s situational judgment (reading a job, sequencing trades, catching trouble early) | Very hard; built over decades | High if near retirement | First |
| Methods your company does differently than the industry standard | Hard; held by a few people | Varies by who holds it | Early |
| Hands-on craft technique learned by doing | Hard to write down, possible to demonstrate | High in aging trades | With video and observation |
| Documented code, specs, and standard procedures | Already on paper | None | Low |
This split between self-performing contractors and the firms managing them matters here. If you run your own crews, the knowledge at the top of that table is craft and trade skill. If you manage sub-trades, it is process, coordination, and the judgment your superintendents and project managers use to keep a job moving. Same risk, different content.
Why “just write it down” falls short on a jobsite
Because the most valuable knowledge is the part the expert can no longer see they have. Ask a master to write down how they do the work and you get the steps they can still narrate. The judgment that makes them a master has gone automatic over the years, and automatic knowledge is close to invisible to the person holding it.
There is a version of this specific to the trades: a lot of the knowledge is physical. A welder can show you what a good bead looks and sounds like. Putting that into words precise enough to teach from is a different and much harder task, and not one most experts can do on demand. So a binder written by your best hand tends to fail in one of two ways. It comes out too thin, because the hard parts felt obvious to the person writing, or it runs so long that no green crew will ever read it.
What does capturing construction knowledge actually involve?
It is a structured process run by someone whose job is to pull the knowledge out, usually built around interviews and, in the trades, observation of the actual work. The expert supplies the expertise; the interviewer supplies the structure and the follow-up questions. It is not a meeting where you hand a superintendent a template and ask them to organize their own career.
In practice the work has a few distinct moves:
- Capture where the work happens: for hand skills, the camera matters as much as the conversation. A recorded walk-through of a setup, a repair, or an inspection captures the parts a transcript misses.
- Interview for the tacit layer: a trained interviewer, usually an instructional designer, keeps pressing the “how did you know that” questions that turn “you just know” into something a newer worker can follow. The interview side of this work is covered further in our guide to eLearning content development.
- Compress when the clock is short: when someone is retiring on a set date, a Knowledge Capture Workshop concentrates the work into a half-day or multi-day intensive that gets the same result on a tighter timeline.
What you walk away with is the raw material a designer can build from: a structured record of the expertise. Turning that into custom training a crew will actually use is the next job, and it is a separate one with its own timeline.
Who does what: your experts or an instructional designer?
Your tradespeople and superintendents are there to demonstrate, explain, and confirm what is accurate. The learning design, the structure, the assessment, and the built course belong to an instructional designer. Blurring that line is how capture projects stall.
What to ask of your expert:
- Whether the content is accurate and current
- The judgment behind the rules: when they bend in the field, and what gets missed
- The stories, near-misses, and mistakes they see most often
- A final check that the finished training is technically right
What an instructional designer handles:
- Sequencing that content into something a newer worker can follow
- Turning field judgment into objectives and realistic scenarios
- Building assessment that checks whether someone can do the task, not recall a slide
- The build, the interactions, and the LMS packaging
The split protects two things at once. An expert pulled into storyboard reviews and quiz-writing is an expert who stops returning your calls, and you cannot afford that when you have only a few of them left. A course built by someone who knows the trade but not how people learn tends to come out accurate and unteachable.
How much of your people’s time does this take, and how do you protect it?
Less than people fear, if it is scoped right, and the scoping is what decides whether it gets finished. Treat your veterans’ time as the scarcest thing in the project, because on a working jobsite it is. A process that assumes a superintendent will “find time somewhere” tends to stall at the second session, and now you have frustrated the person you most needed engaged.
A few habits keep that from happening. Book a small number of focused sessions on specific dates rather than leaving the commitment open-ended; someone who can see the finish line gives you better sessions than someone who suspects this will eat their quarter. Keep the design work off their plate, for the reasons above. And when a retirement date is set, do the capture before it, even if that means compressing everything else, because the retirement date is the one thing on the schedule you cannot move.
There is a human reason these projects succeed or fail, and it is worth naming. In our intake conversations, the experts who lean into a capture process are almost always the ones who feel they are being treated as the source of something valuable, rather than as a data-entry task to be drained. Senior trades-people in particular tend to care that the work outlasts them, and how you frame the ask changes what you get back.
Can you still capture it after someone has left?
Sometimes, but it is a worse position, and that is the argument for not waiting. Once the expert is gone, you are reconstructing from second-hand sources: the people they trained, the record of what went wrong, the work itself. You can rebuild some of it. The highest-judgment material, the part that made the person worth capturing, is the part that degrades fastest once they are no longer there to be asked.
None of this calls for panic about anyone’s retirement. Treat a known retirement date the way you would a permit or a long-lead order: as a planning deadline. The knowledge is recoverable while the person is still reachable. After that, you are guessing at the parts they made look easy.
How Neovation runs this work
Neovation Custom Learning is your full-service, instant L&D capacity, providing expert instructional designers, eLearning developers, and project managers who turn your organization’s raw expertise into interactive, scalable custom training. On knowledge capture, that means running the structured interviews and field sessions so your veteran only has to show up and talk through the work, and handling the design and build so nobody on your side has to learn a new craft to get it done. How involved we get flexes with the project, and cost tracks complexity more than anything else. The comparison that usually matters is to the cost of losing the knowledge entirely, not to a sticker price.
If you already have instructional design capacity in-house, the process in this piece is one you can run yourself; the interview-and-observe approach does not need a partner, only discipline about scope and your experts’ time. A contractor with a single simple topic may be better served keeping it internal or using an off-the-shelf safety course. When the knowledge is complex, the person is leaving soon, or the training has to hold up to real scrutiny in the field, an outside team usually pays for itself. If that is where you are, you can request a quote or see the kind of work we take on, and we will give you an honest read on whether it makes sense before any commitment.
Frequently asked questions
How do you capture a retiring worker's knowledge before they leave?
Through a structured interview and observation process run by someone whose job is to pull the knowledge out, not by asking the worker to write it down. The interviewer, usually an instructional designer, works from a defined scope, asks the follow-up questions that surface the judgment behind the work, and records demonstrations of any hands-on skill. When the retirement date is close, a compressed Knowledge Capture Workshop concentrates the same work into a short intensive. The output is a structured record a designer can build training from.
What construction knowledge should you prioritize capturing?
Start with the knowledge that would be hardest to replace and is closest to leaving. That usually means a retiring expert’s situational judgment, the methods your company does differently than the industry standard, and hand skills newer workers cannot easily pick up. Documented code, specs, and standard procedures are low priority because they are already written down. If you only get a few sessions, spend them on the people whose absence would break something Monday morning.
How do you capture a hands-on skill that someone learned by doing, not from a manual?
You demonstrate and record it, then have an instructional designer build practice around it. Much craft knowledge is physical and cannot be written into words precise enough to teach from, so video of the actual work captures what a transcript misses. The expert shows the task and explains the decisions inside it; the designer turns that into scenarios and assessments that let a newer worker practice. The expert’s job is to demonstrate and verify, not to design the training.
Can you train a skill after the expert has already retired?
Sometimes, but it is harder, slower, and less complete, which is the case for not waiting. Once the person is gone you are reconstructing from the people they trained and the record of what went wrong, and the highest-judgment knowledge is the part that degrades fastest. Some of it is recoverable; the parts the expert made look easy often are not. A known retirement date is best treated as a planning deadline, the same as a permit or a long-lead order.
